A Robust Weight of Evidence Transformation Method for Credit Risk Modeling Using Local Regression

Hui Wang and Shirong Huang ()
Additional contact information
Hui Wang: Federal Home Loan Bank of Atlanta, 1475 Peachtree St NE, Atlanta, GA 30309, USA
Shirong Huang: Federal Home Loan Bank of Atlanta, 1475 Peachtree St NE, Atlanta, GA 30309, USA

Quarterly Journal of Finance (QJF), 2025, vol. 15, issue 02, 1-29

Abstract: The Weight of Evidence (WOE) variable transformation method is widely used in credit risk analysis. This paper introduces a two-staged, local regression-based binning method to estimate WOE. Using a dataset from the banking industry, the study demonstrates that selecting an appropriate number of bins and smoothing factor minimizes information loss and enhances prediction accuracy. The proposed method performs well on imbalanced datasets and can handle both monotonic and U-shaped relationships between the transformed WOE and the original variable, ensuring business soundness. This approach enables smoother credit score migration when financial ratios shift between bins. Given the advantages of the WOE method, such as handling missing values and maintaining model interpretability, the proposed method shows superior performance compared to existing variable transformation approaches, making it highly suitable for financial risk analysis, especially credit risk analysis.

Keywords: Credit scoring models; Bayesian method; weight of evidence; credit risk (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.worldscientific.com/doi/abs/10.1142/S2010139225400026
Access to full text is restricted to subscribers

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:wsi:qjfxxx:v:15:y:2025:i:02:n:s2010139225400026

Ordering information: This journal article can be ordered from

DOI: 10.1142/S2010139225400026

Access Statistics for this article

Quarterly Journal of Finance (QJF) is currently edited by Fernando Zapatero

More articles in Quarterly Journal of Finance (QJF) from World Scientific Publishing Co. Pte. Ltd.
Bibliographic data for series maintained by Tai Tone Lim ().